402
Biochemistry
DON M. CARLSON, ANN L. SWANSON, AND SAUL ROSEMAN
Schubert M. (1947), J. Am. Chem. SOC.69,712. Schwarzenbach, G., and Michaelis, C. (1938), J. Am. Chem. SOC.60,1667. Tarbell, D. S . (1961), in Organic Sulfur Compounds, Kharasch, N.,ed., New York, Pergamon, p. 101.
Troeger, J., and Eggert, A. (1896), J . Prakt. Chem. 53, 482. Weissberger, A., and James, T. H. (1939), J . Am. Chem. SOC.61,442. Weissberger, A., and Snell, J. M. (1939), J. Am. Chem. SOC.61, 450.
Preparation of Crystalline a-u-Galactosamine-1-Phosphoric Acid and Its Conversion to UDP-N-Acetylgalactosamine* DONM. CARL SON,^ ANN L. SWANSON, AND SAUL ROSEMAN
From the Rackham Arthritis Research Unit and the Department of Biological Chemistry, The University of Michigan, Ann Arbor Received October 18, 1963 A simple method has been developed for the enzymatic preparation of a-D-galactosamine-1phosphoric acid in gram quantities, and the product has been crystallized. The compound, obtained in 60% yield, has an [a]*,: = +142.6" (c = 2.0% in HzO). After N-acetylation with acetic anhydride, the resulting N-acetyl-a-D-galactosamine-1-phosphate was crystallized as the potassium salt which had an [ a ] : = $112.4' (c = 2.97, in H2O). UDP-N-acetylgalactosamine was prepared by reacting mono-(tri-n-octylammonium)-N-acetylgalactosamine1-phosphate with uridine-5 '-phosphoromorpholidate in anhydrous pyridine. As the product, UDP-N-acetylgalactosamine, is synthesized by an unambiguous route from the a anomer of galactosamine-1-phosphate, it also must be the LY anomer. A new procedure has been devised for the assay of galactokinase that is generally applicable to enzyme-catalyzed reactions where reducing sugars are converted to glycosides. This assay utilizes sodium borohydride for reduction of excess substrate to the corresponding sugar alcohol. The glycosidic bond, the aglycone being phosphate in the present case, is cleaved with acid, and the resulting sugar is measured by conventional techniques. N-Acetyl-n-galactosamine (2-acetamido-2-deoxy-~galactose) is found in many macromolecular substances (Kent and Whitehouse, 1955); a fungal polysaccharide (Distler and Roseman, 1960) is exceptional in that it contains D-galactosamine as well as the N-acetyl derivative. The precursor of the macromolecules is probably UDP-N-acetylgalactosamine, enzymatically derived from UDP-N-acetylglucosamine (Maley and Maley, 1959; Glaser, 1959). Although both enzymatic and chemical methods can be used for the synthesis of UDP-N-acetylgalactosamine (Davidson and Wheat, 1963; Kim and Davidson, 1963), these procedures are not convenient for preparing pure, specifically labeled material in substrate amounts. Sugar-nucleotides similar to UDP-N-acetylgalactosamine can be synthesized by a general method (Roseman et al., 1961), but only if the corresponding 1phosphate derivative of sugar is readily available. The problem of obtaining galactosamine-1-phosphate in adequate amounts, specifically the cy anomer, was therefore the primary purpose of this investigation. A simple method has been developed for the enzymatic synthesis of D-galactosamine-1-phosphate in gram quantities and, in addition, the pure compound has been crystallized. The hexosamine-phosphate was quantitatively N-acetylated by the usual procedure (Distler et al., 1958), and the N-acetylgalactosamine-lphosphate was converted to the desired UDP-Nacety lgalactosamine. The original procedure for the enzymatic synthesis of galactosamine-1-phosphate (Cardini and Leloir,
* The Rackham Arthritis Research Unit is supported by a grant from the Horace H. Rackham School of Graduate Studies of The University of Michigan. This investigation was supported in part by grants from the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health (A-512), from the American Cancer Society, and from the Michigan Chapter, Arthritis and Rheumatism Foundation. t Postdoctoral Fellow, National Institutes of Health.
1953) was repeated and gave only small amounts of material, a result explained by the observation that yeast galactokinase exhibits a high K, for galactosamine (Alvardo, 1960). By modifying the original conditions and applying methods previously used for isolating the crystalline 6-phosphate esters of glucosamine, galactosamine, and mannosamine (Distler et al., 1958; Jourdian and Roseman, 1962), we have obtained crystalline galactosamine-1-phosphate in better than 60 % yield. Since the enzyme incubations required high galactosamine concentrations, it was not practical to assay the crude reaction mixtures by the usual method, i.e., determining substrate disappearance. A new assay technique was therefore devised that made it possible to measure product formation. In this assay, the excess substrate is reduced with sodium borohydride, to which the product is resistant because it contains a protected carbonyl group. This principle is generally applicable to enzyme reactions where reducing sugars are converted to nonreducing products such as glycosides. The product is measured, in the case of galactosamine-1-P, by N-acetylation, hydrolysis, and a modified Morgan-Elson method. With galactose, the product is measured as reducing sugar after the borohydride treatment, hydrolysis, and deionization. EXPERIMENTAL PROCEDURE
Materials.--.The yeast used in this study was a galactose-adapted strain of Saccharomyces fragilis, purchased from the Sigma Chemical Co. Galactosamine-hydrochloride was prepared as previously described (Roseman and Ludoweig, 1954). Phosphoenolpyruvate, 3-phosphoglyceric acid, UMP, and A T P were commercial materials. UMP-morpholidate was prepared as previously described (Moffatt and Khorana, 1961). Methods. -The following substances were determined by the indicated procedures: protein by the
method of Lowry et al. (1951); N-acetylgalactosamine by a modified Morgan-Elson procedure (Spivak and Roseman, 1959); reducing sugar by the method of Somogyi (1952;) phosphorus by the methods of Telep and Ehrlich (1958) and Fiske and Subbarow (1925). Enzyme Preparation. -A crude enzyme preparation was obtained from galactose-adapted S. fragilis by autolyzing 2 g of yeast in 6 ml of 0.1 M NaHC03 for 12 hours at 25'. After centrifugation at 32,000 X g for 30 minutes the supernatant fluid was decanted. The precipitate was washed with 6 ml of bicarbonate solution, the mixture was centrifuged, and the supernatant fluids were combined. This crude extract contained approximately 50 mg protein per ml. Assay. -The assay mixture contained 200 pmoles of phosphate buffer a t p H 7.8, 2 pmoles of MgCL, 40 pmoles of ATP, 40 pmoles of galactosamine-HCl, 10 pmoles of phosphoenolpyruvate, 10 pmoles of phosphoglyceric acid, and 0.01 ml of the crude galactokinase preparation in a total volume of 0.14 ml. After incubation at 30" for 15 minutes the reaction mixture was heated a t 100" for 2 minutes. Following centrifugation, 0.05 ml of the supernatant fluid was assayed for galactosamine-1-P as described in the following paragraphs. (1) Reduction of the Remaining Substrate with NaBH4.-One drop of capryl alcohol was added followed by 25 pl of 1.0 M NaBH4. After being mixed, the solution was allowed to stand a t room temperature for 5 minutes with occasional shaking, and the borohydride addition was repeated. After 5 minutes 25 p l of 2.0 M acetone was added and again the solution was thoroughly mixed and allowed to stand a t room temperature for 5 minutes. Finally, the excess borohydride was completely destroyed by heating the mixture at 100 " for 3 minutes. (2) Acetylation.-While galactosamine-1-P is extremely resistant to acid hydrolysis (Cardini and Leloir, 1953), the N-acetyl derivative is much more labile (Leloir et al., 1958). The acetylation procedure was that previously described (Distler et al., 1958) and involved the addition of 0.1 ml of 0.5 M aqueous acetic anhydride while the p H was maintained between 7 and 9 with saturated NaHC03. The reaction was complete in about 10 minutes. (3) Hydrolysis.-The N-acetylgalactosamine-1-P was hydrolyzed by adding 0.2 ml of 2 N HC1 followed by heating a t 100' for 10 minutes. After cooling, the reaction mixture was neutralized with 2 N NaOH using phenolphthalein as an indicator. Water was added to a final volume of 1.0 ml and 0.5 ml of the final mixture was assayed directly for N-acetylgalactosamine by the modified Morgan-Elson method. The validity of the assay method was determined by measuring the recovery of added galactosamine-lP. These experiments showed that excess substrate did not interfere with the assay procedure, and that recovery of the product averaged better than 90%. The method can be used for kinetic studies as indicated in Figure 1, which presents the reaction rate as a function of time and protein concentration. A similar assay procedure can be used to follow the phosphorylation of galactose by the crude enzyme preparations. After incubating the enzyme with the substrates as described by Leloir and Trucco j1955), 0.10-ml aliquots of the reaction mixtures were treated with NaBH4 and the borohydride was destroyed by addition of 0.10 ml of 2 N HSO,. The product of enzymatic phosphorylation, galactose-1-P, was hydrolyzed by heating a t 100" for 10 minutes. After cooling, the mixture was neutralized with KHCO,, adjusted to 4 ml with water, and deionized with mixed-
403
GALACTOSAMINE- PHOSPHORIC ACID
Vol. 3, No. 3, March, 1964 I
I
I
I
I
I
W
5
3 v)
0 c
TIME (minutes)
FIG. 1.-Effect of time and protein concentration on galactokinase reaction. Galactosamine was the substrate in incubation mixtures identical to those described under Assay in the text except that all components were increased 10-fold. As shown in the figure, 0.025-0.100 ml of crude extract was added to the incubation mixtures to final volumes of 1.40 ml. Aliquots (0.20 ml) were removed at the indicated times, denatured by heating at 100' for 2 minutes, and assayed as described in the text.
bed ion-exchange resin (Dowex l-HCO3- and Dowex 50, H + ) . Finally, 2-ml aliquots were removed for reducing sugar determinations. Control and recovery experiments with added galactose-1-P showed that this substance was quantitatively determined by the assay procedure. Preparation of Galactosamine-1-phosphate. -A typical incubation mixture for a large-scale preparation of galactosamine-1-P consisted of the following substances in a final volume of 140 ml; 6 mmoles galactosamineHCI; 10 mmoles phosphoenolpyruvate; 6 mmoles ATP; 10 mmoles 3-phosphoglyceric acid; 30 mmoles potassium phosphate buffer, p H 7.8; 3 mmoles MgCL; and 30 ml of the crude galactokinase preparation. The p H was maintained by adding 1 N NaOH as required. After 6 hours a t 30", 0.2 ml of toluene was added and the mixture was maintained at 30" for an additional 12 hours. The reaction was terminated by heating a t 100 for 5 minutes, and the precipitate was removed by centrifugation. Assay of the supernatant fluid showed the presence of 4.8 mmoles of galactosamine-1-P. The solution was placed on a column of Dowex-BO, H + (420 ml, 100-200 mesh) and the column was washed with water to elute the desired product (Jourdian and Roseman, 1962). The product was eluted after the nucleotides and sugar phosphates, and was located by assaying the fractions for total P and galactosamine-1-P. The fractions containing the product were pooled, assayed by the usual procedure, and were found to contain 3.9 mmoles. After lyophilization, 1.1 g of a white powder was obtained, dissolved in 20 ml of water, and treated with 20 ml of 95% ethanol. Crystals slowly formed in the solution, and were harvested after 4-6 days. The first-crop material contained 0.95 g; approximately 0.1 g was obtained in a second crop of crystals. After drying to constant weight over P,Ob in vacuo a t 56", the crystals (Fig. 2) showed the analyses given in Table I. O
1 Elementary analyses were performed by the Spang Microanalytical Laboratory, Ann Arbor, Michigan.
404
Biochemistry
DON M. CARLSON, ANN L. SWANSON, AND SAUL ROSEMAN
TABLE I ELEMENTAL ANALYSES AND OFTICALROTATIONS
Galactosamine-1-P,H1 ~~~
Calculated for
CaH,,OsNP C H N P-
Found 27.67
27.81
5.45 5.41
5.64 5.56 11.88
11.95
UDP-acetylgalactosamine, ~~
Acetylgalactosamine-1-P, ~. Kz Calculated for
CsH,,O.NPK* 25.46 3.74
3.71
Found 25.44 4.10 3.77 8.05
8.21
Lis, 3H,O Calculated for CnHmOmN.P,Li* Found 30.33 4.64
31.00 5.01
6.24
9.76 5.80
9.20
~
[el::
142.6"
(c =
2.0y0in H,O)b
(e
=
112.4' 2 . 9 % in H?O)b
81.7' (c = 0.8% in
H?O).
" None of the compounds contained inorganic P. T h e expected ratios of organic P to hexosamine (as N-acetylgalactosamine) were obtained after acid hydrolysis. The concentrations were based on the organic P content of the solutions. The concentration (pH 7) was calculated from the absorbance at 262 mp, using 10,000 as the extinction coefficient.
FIG. 2.-l'hotom1crograph of galactosamine-1-phosphoric acid crystals. Preparation of N-Acetylgalactosamine-1-phosphate.Crystalline galactosamine-I-P was N-acetylated by the general procedure previously described (Distler et al., 1958). A solution containing 259 mg of galactosamine-I-P, an equivalent of NaHCO,, and 3 ml of methanol in 20 ml of water was cooled in ice. Acetic anhydride (0.47 ml) was added over a 30-minute period with continuous stirring, while the p H was maintained between 6.5 and 8.0 by adding NaHC08. After 1 hour the solution was stirred with 50 ml of Dowex-50, H+, 200-400 mesh, to remove Na+, and filtered, and the resin was washed with water. The acetic acid was removed by repeated extraction with 3 volumes of ether. Control experiments with added "C-acetate showed that six or seven extractions were required for complete removal of acetic acid. The product was converted to its potassium salt hy passage of the solution through a column of Dowex-50 K+, 200-400 mesh. After adjusting the pH to 9.0 with KOH, the product was concentrated in vacuo to a colorless syrup which was insoluble in 95'% ethanol, hut very soluble in methanol. The methanolic solution was treated with a mixture containing equal volumes of ethanol and acetone until a faint persistent turbidity was observed. The addition of the ethanol-acetone mixture was repeated over a period of 3 weeks. The resulting crystals were harvested hy centrifuging, washed twice with 95% ethanol, twice with absolute ethanol, and finally with ether. After drying to constant weight over P,O, in vacuo, the product gave the analyses shown in Table I. While the N-acetylation step was quantitative, the first crop of crystals represented a 35-40y0 yield.
Preparation of Uridine Diphosphate N-Acetylgalactosamine. -The procedure previously described for the synthesis of UDP-N-acetylglucosamine (Roseman et al., 1961) was applied to the preparation of UDP-Nacetylgalactosamine. Thus, 1.0 mmole of the potassium salt of N-acetylgalactosamiue-I-Pwas converted to the his-(tri-n-octylammonium) salt, and this product reacted with uridine-5'-phosphoromorpholidate (0.5 mmole) in anhydrous pyridine (20 ml) for 5 days. The reaction mixture was treated as described and fractionated on Dowex-1, CI- with LiCl in dilute HCI. The yield of purified UDP-N-acetylFalactosamine was 65%; analysisofthissuhstance gave the results shown in Table I. The isolated product was shown to he homogeneous in several of the paper chromatographic and electrophoretic systems used for other sugar nucleotides (Roseman et al., 1961). After heating a t 100" for 10 minutes in 0.1 N HCI, 1.02 moles of N-acetylgalactosamine were released per mole of uridine nucleotide. Preparation of "C-labeled UDP-N-acetylgalaetosamine.*-Galactosamine-I-P was N-acetylated with 14Cacetic anhydride. When this material was treated as described above, without crystallization of the Nacetylgalactosamine-1-P salt, "C-acetyl-UDP-N-acetylgalactosamine was isolated in 50% yield (based on 0.10 mmole of UMP-morpholidate). This product was found to he active in a system that enzymatically synthesized acidmucopolysaccharides (Perlman and Dorfman, 1963). DISCUSSION Anomeric Configuration.--.Galadokinase is known to catalyze the synthesis of the c1 anomer of D-galactoseI-P (Wilkinson, 1949). While Cardini and Leloir (1953) used the kinase to prepare galactosamine-I-P, they did not report the optical rotation of their product. They N-acetylated the compound chemically, however, and reported optical rotations of the N-acetylgalactosamine-I-P, (Cardini and Leloir, 1957). A similar substance was isolated hy direct phosphorylation of N-acetylgaladosamine with an animal kinase (Leloir et al., 1958). The rotations indicated that the phosphate esters were the o( anomers. I n addition, Davidson and Wheat (1963) reported the optical rotation of their preparation of N-acetylgalactosamine-I-P as shown in Table 11. The molar optical rotations reported by the previous workers differ significantly and also disagree with the values observed with our preparation. For purposes of comparison, the molar rotations of D-galactosamine-I-P and its N-acetyl derivative as T h e laheled UDP-A-acetylgalactosamine was prepared by Mr. Jack Distler.
Vol. 3, No. 3, March, 1964
GALACTOSAMINE-1-PHOSPHORIC ACID
405
TABLE I1 OPTICALROTATIONS OF SOMEG GLYCOSIDES AND ~-GLYCOSIDE-~-PHOSPHATES [ a 15
IM 1“
Reference
Glucose-1-P, Kz, 2H20 Glucosamine-1-P, K N-Acetylglucosamine-1-P, K, Methyl D-glucoside Methyl-D-glucosaminide, HC1 Methyl-N-acetylglucosaminide
78 100 79 157 127 105
29,000 29,700 29,800 30,500 29,000 24,600
Krahl and Cori, 1949 Maley et al., 1956 Maley et al., 1956 Hudson, 1909 Neuberger and Rivers, 1939 Neuberger and Rivers, 1939
Galactose-1-P, Ba Galactosamine-1-P, HS M ethyl-D-galactoside
96 142 179
37,900 38,900 38,000
Hansen et al., 1955
Methyl-N-acetylgalactosaminide Methyl-N-acetylgalactosaminide N-Acetylgalactosamine-1-P, KS N-Acetylgalactosamine - 1-P, H 2 N-Acetylgalactosamine- 1-P, Hz N-Acetylgalactosamine-1-P, Li2, 2H20 N-Acetylgalactosamine-1-P,K?
183 187 112 178
44,600 45,600 44,000 54,000 52,000 66,000 46,300
-~
Compound
189 118
b
Dale and Hudson, 1930 b
Masamune et al., 1951 b
Cardini and Leloir, 1957 Leloir et al., 1958 Davidson and Wheat, 1 9 6 3 ~ b
a All rotations are given a t the sodium D line except the last two. Davidson and Wheat report [ a ] * j j ifor s their preparation of N-acetylgalactosamine-1-P. Our sample was therefore also examined a t this wavelength for purposes of comparison (temp. 23’). The [“ID of our preparation was checked with the crystalline K salt and immediately after quantitative N-acetylation of the crystalline galactosamine-1-P (without isolating the product). The values agreed within 3 4r,. b The blank spaces indicate compounds prepared in this laboratory. The methyl a-N-acetyl-D-galactosaminide was prepared according to the method of Stacey (1944) and recrystallized, with large losses until the mp was constant a t 218-220’ (decamp, uncorr). c The concentration of the sugar in solution was determined by analysis for “acid-labile” phosphate according to a personal communication from Dr. E. Davidson. Incomplete hydrolysis would give a high [ c ~ ] 2 6 ~ , ~ .
determined in this laboratory are given in Table I1 along with the values previously reported for Nacetylgalactosamine-1-P and of a variety of a-glycosides. If the first six compounds are regarded as derivatives of a-D-glucopyranose, excellent agreement in the molar rotations is observed, with the possible exception of one case. Thus the correlation is apparently independent of the nature of the aglycone group, methyl or phosphate, or whether the group attached to C-2 is hydroxyl, amino, or acetamido. Similarly, the molar rotation for galactosamine-1-P reported here, 38,900,agrees very well with the values found for the a anomers of galactose-1-P and methyl galactopyranoside, 37,900 and 38,000. While the value of 43,100 obtained with our preparation of Nacetylgalactosamine-1-P shows some disparity with these constants, it agrees with the two values found for methyl-a-N-acetylgaladosaminide, approximately 44,000. These comparisons therefore show that the derivatives reported here are the a anomers. Since the UDP-N-acetylgalactosamine is synthesized by an unambiguous route from the galactosamine-1-P, we conclude that the sugar nucleotide is also the desired a anomer.
REFERENCES Alvardo, F. (1960), Biochim. Biophys. Acta 41, 233. Cardini, C . E., and Leloir, L. F. (1953), Arch. Biochem. Biophys. 45, 55. Cardini. C. E.. and Leloir, L. F. (1957). J . Biol. Chem.225, 317. Dale, J. K., and Hudson, C. S. (1930), J . A m . Chem. Soc. 52, 2534. Davidson, E. A., and Wheat, R. W. (1963), Biochim. Biophys. Acta 72, 112. Distler, J. J., Merrick, J. M., and Roseman, S. (1958), J. Biol. Chem. 230, 497.
Distler, J. J., and Roseman, S. (1960), J . Biol. Chem. 235, 2538. Fiske, C. H., and Subbarow, Y. (1925), J. Biol. Chem. 66, 375. Glaser, L. (1959), J . Biol. Chem. 234, 2801. Hansen, R. G., Rutter, W. J., and Krichevsky, P. (1955), Biochem. Prep. 4, 1. Hudson, C. S. (1909), J . A m . Chem. SOC.31, 66. Jourdian, G. W., and Roseman, S. (1962), Biochem. Prep. 9, 44. Kent, P. W., and Whitehouse, M. W. (1955), Biochemistry of the Amino Sugars, New York, Academic, p. 241. Kim, T. Y., and Davidson, E. A. (1963), Federation Proc. 22, 239. Krahl, M. E., and Cori, C . F. (1949), Biochem. Prep. 1, 33. Leloir, L., Cardini, C. E., and Olavarria, J. M. (1958), Arch. Biochem. Biophys. 74, 84. Leloir, L. F., and Trucco, R. E. (1955), Methods Enzymol. 1 , 290. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951), J . Biol. Chem. 193, 265. Maley, F., and Maley, G. F. (1959), Biochim. Biophys. - Acta 31, 577. Maley, F., Maley, G. F., and Lardy, H. A. (19561, J . A m . Chem. SOC.78, 5305. Masamune, H., Maki, M., and Hiyama, N. (1951), Tohoku J . Exptl. Med. 54, 313. Moffatt, J. G., and Khorana, H. G. (1961), J. A m . Chem. SOC.83, 649.’ Neuberger, A., and Rivers, R. P. (1939), J . Chem. SOC.122. Perlman,. R.,. and Dorfman, A. (1963). Federation Proc. 22. 413. Roseman, S., Distler, J. J., Moffatt, J. G., and Khorana, H. G. (1961), J. A m . Chem. SOC.83, 659. Roseman, S., and Ludoweig, J. (1954), J . A m . Chem. SOC. 76, 301. Somogyi, M. (1952), J . Biol. Chem. 195, 19. Spivak, C. T., and Roseman, S. (1959),J.A m . Chem. SOC.81, 2403. Stacey, J. (1944), J . Chem. SOC.,272. Telep, G., and Ehrlich, R. (1958), Anal. Chem. 30, 1146. Wilkinson, J. F. (1949), Biochem. J. 44, 460.